Depression is a prevalent condition in the general U.S. population, estimated to impact 6.6% of Americans within a given 12-month observation window (1). To date, information on depression after living kidney donation is largely drawn from single-center surveys and interview-based studies. A 2006 systematic review by Clemens et al. (2) combining data for more than 5000 living donors estimated that depression affects 5% to 23% of donors at an average of four years after nephrectomy. While quality-of-life-scores that are statistically similar to those of controls have been identified in some prior donor samples (3–5), increased depression rates and exacerbations of preexisting depression after donor nephrectomy have also been reported (6, 7).
Study findings have varied in how recipient outcomes impact donors psychologically. A small study of 32 living donors surveyed before and at an average of 19 months after donation found a significant decrease in depression index scores in the cases of successful transplantations (recipient alive with graft function at assessment) (8). In contrast, evidence from case reports and focused interviews demonstrated links between recipient graft failure or death and the onset of psychological symptoms in donors including depression, “mental paralysis,” “mentally falling apart,” and even suicide (9,10). Research teams conducting larger surveys suggested that recipient death or graft failure predicted regret about donation and a lower donor quality of life (4, 11), but a recent multinational survey reported similar quality-of-life scores between donors whose recipients experienced an adverse outcome and donors whose recipients were alive with graft function (5).
Surveys and interviews can yield important information about psychological health outcomes, but these designs also may be limited by sampling and response biases. Administrative billing claims track records of provider-reported diagnoses that offer descriptions of conditions identified in the course of clinical practice. To advance the understanding of depression among a national sample of prior living kidney donors in the United States, we examined a unique database that links Organ Procurement and Transplantation Network (OPTN) identifiers of living kidney donors to billing claims of a private health insurer. Provider-reported depression diagnoses were used to examine the burden and correlates of depression among prior donors, with particular attention to associations of recipient clinical outcomes with donor depression.
Demographic and Clinical Correlates of Depression Diagnoses Among Prior Living Donors
Among 4650 prior living donors in the full study sample, 54.6% were women, and 76.3% were white. Eighty-one percent of donors were biologically related to their recipient, 7.6% were spouses, and 11.2% were nonspousal unrelated donors (Table 1). The median durations from donation to the start and end of captured insurance benefits were 4.9 and 7.7 years, respectively. After the start of the benefits, the cumulative frequency of depression diagnosis was 4.2% at 1 year and 11.5% at 5 years. Among the donors with depression diagnoses, 83.4% filled a prescription for an antidepressant medication at some point within the available pharmacy claims data, with a median time of 9 days between depression diagnosis and prescription fill dates.
In bivariate analysis, clinical correlates of depression diagnoses among the prior donors included donor age, gender, and race (Table 1). Specifically, the likelihood of depression diagnosis rose significantly by 2% with each increase in year of donor age at the start of insurance benefits (adjusted hazard ratio [aHR], 1.02; 95% confidence interval [CI], 1.01–1.03). Depression after the start of insurance benefits was more than twice as likely among female compared with male prior donors (aHR, 2.41; 95% CI, 1.89–3.06) and among white compared with nonwhite donors (aHR, 2.07; 95% CI, 1.50–2.86). In absolute terms, by 5 years after the start of insurance enrollment, claims for depression diagnoses occurred in 7% of male and 15.5% of female prior donors, respectively (P<0.0001). Depression diagnoses were identified among 6.5% of nonwhite and 13.2% of white prior donors, respectively, by 5 years after the start of plan benefits (P<0.0001). Similar bivariate associations with donor age, gender, and race were observed using depression diagnoses with pharmacological treatment as the outcome measure (see Appendix 1, SDC 1, http://links.lww.com/TP/A667).
The likelihood of depression diagnosis among prior donors within the captured insurance data was not significantly associated with the type of donor-recipient relationship or the duration of time from nephrectomy to the start of enrollment in the insurance plan (Table 1). By 5 years after enrollment, the cumulative frequency of depression among biologically related, spousal, and nonspousal unrelated donors was 11.0%, 12.5%, and 16.5%, respectively, but these differences were not statistically significant (P>0.05).
The linked data captured 2264 persons who donated after the start of the OPTN collection of information on surgery type, of whom 48.7% underwent donor nephrectomy by laparoscopic techniques including hand-assisted laparoscopy. Surgery type was not associated with the likelihood of postdonation depression (P=0.31). The sample included 460 persons who donated during the period of the OPTN collection of information on perioperative complications, of whom 0.9% required reoperation, 1.1% showed other complications (including the need for percutaneous procedures), 1.8% required readmission, and 5.4% demonstrated any reported complication (Table 1). Although these events were infrequent, reoperation was strongly associated with postdonation depression diagnoses (aHR, 7.73; 95% CI, 1.05–58.33; P=0.04) in bivariate analyses, and other complications showed a nearly significant association (aHR, 4.27; 95% CI, 0.98–18.48; P=0.05). The association of reoperation with postdonation depression remained significant after adjustment for baseline donor demographic factors (aHR, 9.73; 95% CI, 1.02–99.15; P=0.0 cological treatment as the outcome measure, such that reoperation and other complications were significantly associated with pharmacologically-treated depression after adjustment for baseline demographic traits.
Depression Burden Among Living Donors Compared With Nondonor Controls
In matched-pairs comparisons, depression diagnoses occurred at a rate of 5.0 per 100 person-years in the living-donor sample compared with 7.1 per 100 person-years among age- and gender-matched nondonor beneficiaries in the insurance plan (rate ratio, 0.70; 95% CI, 0.61–0.80) (Table 2). Although depression diagnoses were more common among women than men, donor-to-control rate ratios were similar across genders. Point estimates for depression diagnosis rate ratios in the living donors compared with age- and gender-matched nondonors were less than 1.0 across donor relationship categories (rate ratios, 0.68–0.85), although the confidence interval excluded 1.0 only in the largest donor relationship category of biologically related. We also stratified the living-donor sample according to the time from donation to insurance enrollment as above or below the sample median. The rate of depression diagnoses in living donors (and corresponding donor-to-control rate ratios) was the same regardless of the timing of insurance enrollment (Table 2).
Associations of Recipient Clinical Events With Subsequent Donor Depression
Before donor enrollment in the study insurance plan, 11.4% of donors had lost recipients to death, and 11.7% of donated grafts had failed for other reasons. By five years after donor insurance enrollment, recipient death and graft failure affected another 12.0% and 8.7%, respectively (Kaplan-Meier estimates). After adjustment for baseline donor demographic factors, recipient clinical outcomes predicted approximately twofold to threefold increases in the relative risks of depression among nonspousal unrelated living donors (Table 3A). Specifically, recipient death was associated with a 123% increase in the relative risk of subsequent depression diagnosis (aHR, 2.23; 95% CI, 1.11–4.48) among nonspousal unrelated donors. Recipient death-censored graft loss was associated with a 230% increase in the relative risk of subsequent depression diagnosis (aHR, 3.30; 95% CI, 1.49–7.34) among nonspousal unrelated donors. There were trends toward increased depression risk after recipient death and graft failure among spousal donors (aHR point estimates ≥2.0), but these associations were not statistically significant. Recipient death (aHR, 1.11; P=0.5) and graft failure (aHR, 1.06; P=0.8) did not show evidence of associations with the likelihood of subsequent depression diagnosis among related donors. Consistent findings were observed using depression with pharmacological treatment as the outcome measure (Table 3B).
Psychological outcomes after kidney donation are not systematically tracked in national transplant registries. We examined a database integrating U.S. living-donor registration information with billing claims from a private health insurer to study depression diagnoses in a large national sample of prior donors. The study revealed, first, that depression diagnoses occurred in some prior donors, at an estimated rate of approximately 5.0 per 100 person-years. When compared with an age- and sex-matched sample of general insurance beneficiaries who were not living donors, however, depression diagnoses were less frequent among prior donors versus nondonors in the same insurance system. Therefore, a “red flag” of greater depression burden among donors was not demonstrated. Second, the frequency of depression diagnoses in prior living donors varied by demographic and clinical factors, being more common in women compared with men, in white compared with nonwhite living donors, and among donors who experienced certain perioperative complications. Third, recipient death and graft loss predicted an increased risk of depression diagnosis in unrelated living donors, but there was no association between these adverse recipient outcomes and donor depression for biologically related donors.
Depression is not uncommon in the general population, and thus, one important question is whether kidney donation increases the risk of depression. We found that the rate of depression diagnoses among prior donors was significantly lower than among age- and gender-matched general insurance beneficiaries in the same insurance system. Because donors and nondonors are different individuals facing different physical and psychological health stressors, and donors undergo selection before donation, identifying ideal controls is challenging. Consistent with our use of general controls and our findings, prior studies reported superior quality-of-life and health summary scores among prior donors compared with general U.S. population norms (4, 12). A recent multinational survey found similar quality-of-life scores among donors and controls screened for baseline good health (5).
Although past findings on donor quality of life and the overall low frequency of depression diagnoses observed in the current study are encouraging, some donors will experience feelings of loss, depression, and reduced mental energy (3, 13). Being a donor often means recovering from surgery while serving as a caregiver, making convalescence stressful (3). Even when this period is over, the relief can result in a paradoxical depression (14). Thus, it remains important to identify baseline traits and clinical events that may predict poor psychological outcomes.
Gender-based variation in the frequency of depression is well established in the general population. For example, a nationally representative face-to-face household survey in 48 states estimated that the lifetime odds of depression in American women are 1.7 times the odds in men (95% CI, 1.5–2.0) (1). Similarly, we found that depression diagnoses were more than twice as common among female compared with male adult prior donors and that this pattern was consistent among nondonor beneficiaries of the same insurer. With this clearly established gender difference, it may be that female donors need additional depression screening before donation and psychological support afterward if depression emerges during follow-up.
We also found that white donors were twice as likely to have post-donation depression diagnoses compared with nonwhite donors. Significantly lower lifetime prevalences of major depressive disorder were identified among African Americans (adjusted odds ratio, 0.51) and Hispanics (adjusted odds ratio, 0.38) compared with white persons, in the National Health and Nutrition Examination Survey III (15). The mechanism of lower frequencies of clinically detected depression among persons of other racial and ethnic backgrounds compared with white persons seems to be in part because of minorities’ reluctance in seeking mental healthcare and indirect or “culture bound” presentations. In an investigation of the symptoms of depressed Chinese Americans, Yeung et al. (16) found that 42% of those with major depression initially presented with physical complaints, although none of the subjects reported complaints of mood alteration. Manifestation of depression with somatic rather than mental or emotional symptoms has been identified in Hispanic and Asian Indian populations (17, 18). Future research should investigate whether demographic differences in depression diagnoses after kidney donation reflect care seeking, clinical ascertainment, or true differences in disease burden.
As collection of perioperative complications data by the OPTN began recently, we examined associations of early complications with depression in a subgroup of recent donors. Although complications were infrequent, we observed significant associations of the requirement for reoperation with postdonation depression and associations of other complications including percutaneous interventions with pharmacologically-treated depression. Although the effect estimates are imprecise (i.e., have wide confidence limits) because of the small subsample and rarity of perioperative complications, the results suggest that live donors who experience surgical complications may need focused follow-up of their psychological health. Further study is needed to refine these preliminary findings.
Limited data are available on whether donor-recipient relationship impacts or modifies postdonation psychological outcomes. In a survey of 524 responding prior donors at the University of Minnesota from 1984 to 1996 that was mostly composed of living related donors (92%), a relationship other than first-degree blood relative was associated with a nearly significant trend toward increased odds of regretting donation or viewing the overall experience as very stressful (odds ratio, 3.5; P=0.06) (4). Another survey of 167 donors found that the emotional experience was more positive than expected among 51% of first-degree relatives compared with 31% of donors with more distant relationships (19). In the current study, we found a trend towards more common depression among unrelated compared with biologically related donors, but this pattern was not statistically significant.
Although donor-recipient relationship itself was not a significant correlate of postdonation depression in our study, we found that relationship modified the impact of adverse recipient events on the risk of depression diagnosis in the donor. Specifically, recipient death and graft loss predicted a twofold to threefold increased risk of subsequent depression among unrelated nonspousal donors. To our knowledge, such variations have not been previously examined, but prior studies suggested that adverse recipient events may have psychological consequences for some living donors. A multinational cohort of 203 donors from 1970 to 2007 assessed between 2004 and 2008 (median, 5.5 years after transplantation) reported similar quality-of-life scores in donors whose recipients experienced an adverse outcome and those donors whose recipients were alive with graft function (5). In contrast, recipient death within the first year of transplantation was associated with three times the likelihood of regretting donation or finding the experience very stressful (odds ratio, 3.3; P=0.1) in the University of Minnesota survey (4). A survey that included the responses of 39 donors whose recipients died or lost their graft reported that 43% believed their role as a donor made the failure more devastating and 11% had experienced suicidal ideation (19). In another small survey, 50% of donors believed that their experience was not “worth it” when the recipient died (20). Two donor suicides after graft loss or death of the recipient were examined as case reports (10).
Our observation of the varying impacts of adverse recipient outcomes according to donor-recipient relationship is hypothesis generating but cannot be explained with the available data. One hypothesis is that nonrelated and, thus, more altruistic donors are more psychologically vulnerable to failure of their gift, but further study is needed to assess the impact of recipient events on donor depression using a broader array of measures. At present, we agree with the recommendations of Litwin (21) and Weizer et al. (10) that transplantation teams should attempt to assess and attend to the mental health of all donors after recipient death or graft failure, even if the transplantation was years ago. If the donor’s care has transitioned to a primary care provider, this may require communication and assessment through the current provider.
Limitations of this study include factors related to the sample and outcome measures. The claims data are derived from a private insurance plan, and uninsured living donors are not captured. Claims are surrogate measures for diagnoses, and coding errors are possible. Clinical recognition of depression in particular is impacted by the willingness of the patient to seek care and directly present symptoms (15, 17, 18), and because depression codes may be differently interpreted by clinicians, claims do not offer the severity information that can be obtained through surveys and interviews (22). The available data limit the capture of diagnoses to benefits periods of the linked health insurance, and thus, we focused on diagnoses during available observation windows of benefits under the insurer, which began at a median of 5 years after donation. Predonation benefits were captured for only a minority of the donors (14.4%), and thus, information on predonation depression diagnoses was not available for inclusion. Of note, the depression diagnosis rate among donors was similar regardless of time from nephrectomy to plan enrollment in this analysis. Further, given that psychological outcomes are not systematically tracked after donation, even large survey studies frequently use designs that capture donors at varying times after donation (4, 5). Given the duration of available insurance benefits, we did not have sufficient data to screen controls, and thus, the donor-control comparisons must be interpreted with the limitations of other general comparisons (4, 12).
In conclusion, in this privately insured sample of prior donors, we found that depression diagnoses were more common in women, in white donors, and in unrelated donors after recipient graft failure or death. Although the study design and available data are inherently impacted by care seeking patterns and patient-provider communication, these data offer an additional perspective on postdonation mental health outcomes diagnosed in real practice that supplement information obtained through surveys, interviews, and case reports. Additional studies should investigate whether demographic differences in depression diagnoses after kidney donation reflect care seeking, clinical ascertainment, or true differences in disease burden. Although most donors seem to experience good or stable psychological health (2, 4, 5), it is incumbent on the transplantation community to seek to identify those at risk for depression and other mental illnesses. At this time, we believe that informed consent and postdonation care should attend to the potential impact of perioperative complications and recipient outcomes on the donor’s psychological health.
MATERIALS AND METHODS
Data Source and Sample
Study data were assembled by linking OPTN records for prior living kidney donors with administrative data from a national private health insurer. OPTN records include information on all donors and transplant recipients in the United States as submitted by OPTN member centers. After the approval of the Health Resources and Services Administration and the Saint Louis University institutional review board, beneficiary identifier numbers from the insurer’s electronic databases were linked using names and birthdates to unique OPTN identifiers for living kidney donors. Analyses were performed using Health Information Portability and Accountability Act–compliant limited data sets with all direct identifiers removed.
People were eligible if they had OPTN records of serving as a living kidney donor sometime between October 1987 and July 2007 and benefits under the participating insurer after donor nephrectomy at some point from May 2000 to December 2007 (the period of available claims data). All study participants were simultaneously enrolled in medical and pharmacy benefits with this insurer exclusively during the study window. Because of the large sample size, the anonymity of the patients studied, and the nonintrusive nature of the research, a waiver of informed consent was granted per the Department of Health and Human Services Code of Federal Regulations (Title 45, Part 46, Paragraph 46.116).
Definitions of Outcomes and Covariates
Living-Donor Characteristics and Recipient Clinical Events
Demographic data from the OPTN at donor nephrectomy included age, gender, race, and donor-recipient relationship status as reported by the transplantation center. Time from donation to the start of captured insurance benefits (at or after donation) and age at the start of benefits were computed for use in “time-to-depression diagnosis” models. The relationship of the donor to the recipient was defined by the OPTN report as biologically related, spouse or partner, or other. Recipient graft failure and associated dates were drawn from OPTN reports based on an indication of the permanent return of the transplant patient to dialysis or retransplantation. Recipient death events were identified by OPTN reporting and supplemented with the Social Security Death Master File. Because OPTN collection of descriptive information on the donation procedure has evolved over time, we performed subanalyses considering the type of donor nephrectomy surgery (open vs. laparoscopic, collection starting 1998) and perioperative complications reported to the OPTN within 6 weeks of donation (collection starting 2004) among subsamples who donated during the period of collection of the relevant variables.
Depression Diagnoses and At-Risk Periods
Depression diagnoses were defined by identification of billing claims with a corresponding diagnosis (International Classification of Diseases Code Ninth Revision, Clinical Modification) code for major depression or depression not otherwise specified (296.2, 296.3, 311) in the primary analyses. In secondary analyses, we defined depression as a provider-reported depression diagnosis on a billing claim with concomitant or subsequent pharmacy claims for antidepressant medications (see Appendix 2, Supplemental Digital Content 1, http://links.lww.com/TP/A667). A schematic of the study design, constructed to assess associations of donor demographic traits and recipient clinical outcomes with the burden of depression in this sample, is shown in Figure 1. The available data captured insurance enrollment windows for health outcomes assessment at varying times in relation to nephrectomy across the study sample. Time-to-depression diagnosis was computed as the time from the start of captured postdonation benefits to submission of the first depression claim, adjusted for the time from nephrectomy to the start of benefits. At-risk time was right censored at the end of the insurance benefits or December 2007.
Correlates of Depression After Donation
Data sets were merged and analyzed with SAS for Windows software, version 9.2 (SAS Institute Inc., Cary, NC). Bivariate associations (hazard ratio) of factors reported before the index date (start of captured postdonation benefits), including donor demographics, time from donation to the index date, and donor-recipient relationship, were assessed by Cox regression. Cox regression was also used to estimate a probability distribution function for depression diagnosis after the start of benefits under the insurer, adjusted for time from nephrectomy to enrollment in the insurance plan.
Associations (aHR) of recipient posttransplant graft failure and death with donor depression were examined by time-varying Cox regression. The time-varying regression considers the recipient events that occurred before donor insurance enrollment and postenrollment recipient events as potential predictors of subsequent donor depression diagnoses. Recipient graft failure and death events were further characterized by the donor-recipient relationship. Stratification by relationship was used to establish donors in a given relationship category who remained without depression diagnosis as reference groups. This allowed estimation of all time-dependent variables within a single model, facilitating comparison of effect estimates, as illustrated in previous studies (23). The time-varying covariates were included in all final models, and stepwise selection was performed across other factors with a P value for retention of less than 0.05.
Comparison of Depression Frequency in Donors and Nondonor Controls
To compare the burden of depression among the donors to a nondonor sample, donors were matched one-to-one with general insurance beneficiaries (nondonors) by gender and age when benefits began. The maximum observation time defined by benefits duration in each matched pair was limited to the shortest available in the pair. We compared the number of patients with depression per observation time among the donors versus age- and gender-matched controls as rate ratios. Comparisons were stratified according to donor traits of interest.
1. Kessler RC, Berglund P, Demler O, et al.. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). Jama 2003; 289 (23): 3095.
2. Clemens KK, Thiessen-Philbrook H, Parikh CR, et al.. Psychosocial health of living kidney donors: a systematic review. Am J Transplant 2006; 6 (12): 2965.
3. Andersen MH, Bruserud F, Mathisen L, et al.. Follow-up interviews of 12 living kidney donors one yr after open donor nephrectomy. Clin Transplant 2007; 21 (6): 702.
4. Johnson EM, Anderson JK, Jacobs C, et al.. Long-term follow-up of living kidney donors: quality of life after donation. Transplantation 1999; 67 (5): 717.
5. Clemens K, Boudville N, Dew MA, et al.. The long-term quality of life of living kidney donors: a multicenter cohort study. Am J Transplant 2011; 11 (3): 463.
6. Minz M, Udgiri N, Sharma A, et al.. Prospective psychosocial evaluation of related kidney donors: Indian perspective. Transplant Proc 2005; 37 (5): 2001.
7. Prandini R, Bonomini V, Vangelista A, et al.. Living donors
in renal transplantation: a long-term study. Transplant Proc 1987; 19 (1 pt 2): 1498.
8. Frade IC, Fonseca I, Dias L, et al.. Impact assessment in living kidney donation: psychosocial aspects in the donor. Transplant Proc 2008; 40 (3): 677.
9. Haljamae U, Nyberg G, Sjostrom B. Remaining experiences of living kidney donors more than 3 yr after early recipient graft loss. Clin Transplant 2003; 17 (6): 503.
10. Weizer N, Weizman A, Shapira Z, et al.. Suicide by related kidney donors following the recipients’ death. Psychother Psychosom 1989; 51 (4): 216.
11. Giessing M, Reuter S, Schonberger B, et al.. Quality of life of living kidney donors in Germany: a survey with the Validated Short Form-36 and Giessen Subjective Complaints List-24 questionnaires. Transplantation 2004; 78 (6): 864.
12. Ibrahim HN, Foley R, Tan L, et al.. Long-term consequences of kidney donation. N Engl J Med 2009; 360 (5): 459.
13. Taghavi R, Mahdavi R, Toufani H. The psychological effects of kidney donation on living kidney donors (related and unrelated). Transplant Proc 2001; 33 (5): 2636.
14. Fukunishi I, Shindo M, Ohara T, et al.. “Paradoxical depression
” in a living female donor after kidney transplantation
. Psychosomatics 1998; 39 (4): 396.
15. Riolo SA, Nguyen TA, Greden JF, et al.. Prevalence of depression
by race/ethnicity: findings from the National Health and Nutrition Examination Survey III. Am J Public Health 2005; 95 (6): 998.
16. Yeung A, Chan R, Mischoulon D, et al.. Prevalence of major depressive disorder among Chinese-Americans in primary care. Gen Hosp Psychiatry 2004; 26 (1): 24.
17. Raguram R, Weiss MG, Channabasavanna SM, et al.. Stigma, depression
, and somatization in South India. Am J Psychiatry 1996; 153 (8): 1043.
18. Mischoulon D. Psychiatric management of Hispanic patients: cross-cultural issues and ethnopsychopharmacology. Program and Abstracts of the American Psychiatric Association 2004 Annual Meeting 2004: Symposium 75C.
19. Schover LR, Streem SB, Boparai N, et al.. The psychosocial impact of donating a kidney: long-term follow-up from a urology based center. J Urol 1997; 157 (5): 1596.
20. Brown CJ, Sussman M. A transplant donor follow-up study. Dial Transplant 1982; 11: 897.
21. Litwin MS. Editorial comment: the psychosocial impact of donating a kidney. J Urol 1997; 157: 1600.
22. Melfi CA, Croghan TW. Use of claims data for research on treatment and outcomes of depression
care. Med Care 1999; 37 (4 Suppl Lilly): AS77.
23. Lentine KL, Xiao H, Brennan DC, et al.. The impact of kidney transplantation
on heart failure risk varies with candidate body mass index. Am Heart J 2009; 158 (6): 972.